1. Introduction

Prostate cancer (PCa) is the most prevalent cancer among men and the second most diagnosed worldwide. While initial responses to conventional therapies are often positive, advanced PCa frequently leads to metastasis and therapy resistance. Immune checkpoint blockade (ICB) has shown promise in treating tumors; however, the immunosuppressive tumor microenvironment (TME) in PCa limits its efficacy. In this sense, epigenetic inhibitors have demonstrated potential in reprogramming immune cells towards a tumoricidal phenotype. The analysis of The Cancer Genome Atlas (TCGA) databases revealed significant alterations in RNA modifying proteins (RMP), known to influence tumor progression. Applying deconvolution methods, we found an inverse correlation between RMP expression and immune cell infiltration. Despite this, no RMP signature has been directly linked to PCa advancement. In this study, we aim to elucidate the role of RMPs in PCa progression and their interaction with immune cells.

2. Experimental design

Prostate tissue was extracted from four different mouse models:

  • Healthy: healthy prostate tissue.

  • Healthy_Cast: healthy tissue from a castrated mouse.

  • Tumour: tumoral prostate tissue.

  • Tumour_Cast: tumoral tissue from a castrated mouse (castration-resistant mouse).

3. Results

Stage dependent cellular dynamics

Unsupervised graph-based clustering of all samples visualized by UMAP, delineated by cell type.

Stacked bar indicating the percentage abundance of each cell type in each sample.

UMAP of all cells (left) or epithelial cells (right) separated by sample. The black circle highlights the epithelial population, while the red circle surrounds a specific epithelial subcluster that significantly varies across samples. Pbsn is a typical prostate marker, and its promoter was used to selectively delete Pten.

Epithelial cells subclustering

Unsupervised graph-based subclustering of the epithelial populations (luminal, basal and periurethral).

Stackbar representing percentage abundance in each sample.

Characterization of different subpopulations in terms of gene expression markers
Characterization of different subpopulations in terms of gene expression markers


Pseudotime analysis in each sample. Cells with higher pseudotime values are more differentiated from the origin populations. The red asterisk (*) indicates the suggested origin cells.
Pseudotime analysis in each sample. Cells with higher pseudotime values are more differentiated from the origin populations. The red asterisk (*) indicates the suggested origin cells.

Increased immune interaction and migration in epithelial subpopulations after castration

Gene Ontology (GO) enrichment analysis of upregulated biological processes in epithelial subpopulations 7 and 8 in healthy and castrated resistant tissues.

Increased epithelial-immune system interaaction in tumoral tissue after castration

Chord plot showing the differential number and strength of interactions among different cell populations across two samples.

Differential expression of RNA modifyng proteins across different epithelial subpopulations

Heatmap showing log2 fold-change expression of RNA epitranscriptomic genes in different epithelial subclusters, comparing tumor vs.Ā healthy and tumor castrated vs.Ā tumor.

PCIF1 expression patterns in a spatial tumoral context

UMAP showing Pcif1 expression in epithelial cells.

Multiplex imaging analysis showing coexpression of different immune cells and PCIF1 in each sample. Scale bar represents 50 μm.

4. Conclusions

  • An increased abundance of immune cell populations is observed in tumor and castration-resistant tissues compared to healthy tissue.

  • Castration-dependent and castration-independent epithelial subpopulations have been identified and characterized.

  • Pseudotime analysis based on differentially expressed genes suggests potential origin cells and provides insights into the differentiation trajectory of these cells.

  • Specific subpopulations with stemness and proliferation profiles show enrichment of pathways related to immune system interaction and migration in the castration-resistant context compared to healthy tissue.

  • Cell interaction analysis reveals a higher number of interactions between all cells in tumor and post-castration tissues, with epithelial cells interacting more with the tumor microenvironment in the castration-resistant context than in primary tumor tissue.

  • An analysis of a curated list of epitranscriptomic markers highlights RNA-modifying proteins (RMPs) that are specifically overexpressed in tumor and castration-resistant tissues.

  • PCIF1 is differentially expressed in specific epithelial subclusters, showing a wide range of expression within tumor cells. Its correlation with immune cell infiltration is currently being investigated at the spatial resolution level through multiplex imaging.